potentially confounding variables, both known and unknown, allows investigators to reach conclusions that are applicable to individuals and generalized to populations. For this reason, the Panel agrees that RCTs, which address specific questions on the management of clinically localized prostate cancer, deserve special consideration.
RCTs were identified from the pool of articles generated by the Guideline Panel and from the Cochrane trials registry for prostate cancer, which was last updated on September 2, 2005. Articles selected for discussion herein were limited to studies executed as prospective RCTs that investigated the impact of interventions on treatment outcomes for localized prostate cancer. Some studies culled from the Cochrane registry did not meet the strict criteria established by the Panel but were felt to merit discussion as they provided the best available quality of evidence to answer specific research questions. These limits yielded 27 studies for incorporation into this portion of the Guideline (Tables 1-4).8-13, 15-29, 31-35, 44, 119
Two broad conclusions can be drawn from the review of RCTs for localized prostate cancer and will subsequently be discussed in greater detail. First, there are very few trials investigating a direct comparison of two different treatment modalities (e.g., active surveillance vs. external beam radiotherapy or external beam radiotherapy vs. radical prostatectomy). Second, there are
many RCTs that investigate interventions within a particular treatment modality (e.g., radical
prostatectomy alone vs. neoadjuvant androgen deprivation plus radical prostatectomy or different doses of radiation). As a consequence, the highest quality evidence to identify a superior
treatment modality for a particular patient is lacking, but there is some high-quality evidence to support various modifications within treatment modalities.
RCTs Comparing Different Treatment Modalities Watchful Waiting Versus Radical Prostatectomy
Given the slow progression of many localized prostate cancers, it has long been recognized that not all cases warrant intervention. Two RCTs, one in the pre-PSA era, have reported long-term follow-up of patients randomized to watchful waiting or radical prostatectomy, but the second one is not yet mature. The Veterans Administration Cooperative Urological Research Group
(Table 1)20 reported on 142 patients with clinical stage I or II adenocarcinoma of the prostate
This study was underpowered to detect treatment differences, and applicability of these findings to contemporary patients is limited given both stage and grade migration since the advent of PSA screening for prostate cancer.
More recently, the Scandinavian Prostate Cancer Group Study No. 4 (Table 1)10 reported on 695
men with clinical stage T1 or T2 prostatic adenocarcinoma (comparable to current T1 to T2N0M TNM stage) who were randomized to watchful waiting (n=348) or radical prostatectomy (n=347) between 1989 and 1999. Although this trial was conducted after PSA level testing was available, only 5% of men were diagnosed by screening. Still, the distribution of serum PSA levels at the time of diagnosis more closely reflects contemporary populations in which PSA screening is widespread. After a median follow-up of 8.2 years, treatment with radical prostatectomy was associated with significantly lower risk of disease-specific mortality, overall mortality, metastatic disease, and local progression (Table 5).10
Table 5. Outcomes of the Scandinavian Prostate Cancer Group Study No. 4: median follow-up of 8.2 years10
RP % (n)
WW
% (n) Relative risk (95% CI) p value
Numbers needed to treat Disease-specific mortality 9.6% (30) 14.9 % (50) 0.56 (0.36 to 0.88) 0.01 20 Overall mortality 27% (83) 32% (106) 0.74 (0.56 to 0.99) 0.04 20 Distant metastasis 15.2% (50) 25.4% (79) 0.60 (0.42 to 0.86) 0.004 10 Local progression 19.2% (64) 44.3% (149) 0.33 (0.25 to 0.44) <0.001 4 CI, confidence interval;RP, radical prostatectomy; WW, watchful waiting.
In preplanned subset analyses, the investigators found that the reduction in risk of death from prostate cancer in those randomized to prostatectomy was more pronounced in the population of men less than 65 years of age and independent of PSA level or Gleason score at diagnosis (p=0.08 for treatment by age-group interaction). However, caution must be used in interpreting subset analyses.
The Prostate Cancer Intervention Versus Observation Trial (PIVOT)50 is an ongoing RCT comparing radical prostatectomy to watchful waiting in patients with clinical stage T1 or T2 disease. Initiated in 1994, accrual was slow and finally was completed with an enrollment of 731 patients in 2002. Follow-up is planned for 15 years, with overall mortality as the primary
endpoint. Although findings will not be available for some time, study findings will be more applicable to contemporary patients diagnosed with localized prostate cancer.
Adjuvant Bicalutamide Therapy
The bicalutamide Early Prostate Cancer Program was a multicenter series of three international RCTs launched to assess the efficacy and tolerability of bicalutamide, either alone or in
combination with radical prostatectomy, radiation therapy, or watchful waiting, in patients with clinically localized or locally advanced prostate cancer. Approximately two thirds of the patients had localized disease. This program included three separate controlled trials designed to allow for combined analysis (Table 1).20, 33, 119 The North American trial119 included patients who
mainly opted for prostatectomy, the trial conducted in Europe33 and other countries worldwide
enrolled primarily patients receiving radiotherapy, and the Scandinavian study20 was comprised
primarily of patients choosing watchful waiting. Each study had similar endpoints, but
bicalutamide treatment duration differed across the three studies. Early reports and a subsequent
analysis with longer follow-up33 have consistently demonstrated significantly improved
progression-free survival with bicalutamide in the overall study population compared to placebo, but no overall survival benefit was seen. A number of subset analyses were performed based on study number, primary treatment received, clinical stage, and other factors. One analysis
conducted at a median of just over five years of follow-up indicated that men with localized prostate cancer managed with watchful waiting plus bicalutamide had reduced overall survival in
comparison to men managed with watchful waiting alone.20, 33 Because the risk of a false-
positive result increases with multiple statistical testing, this must be considered when evaluating the results of subset analyses. While the explanation for this difference in overall survival noted in this subgroup analysis is not readily apparent, there is some suggestion that men who are considering watchful waiting for their clinically localized prostate cancer may not benefit from the addition of bicalutamide as part of their immediate therapy.
RCTs Within Treatment Modalities External Beam Radiotherapy
External beam radiotherapy dosage. Three recent RCTs have compared different external beam
radiotherapy dosages. The first, from M. D. Anderson Hospital (Table 2),27 compared the
efficacy of 70 versus 78 Gy in 305 patients with clinical stage T1 to T3N0 prostate cancer randomized between 1993 and 1998. The primary endpoint was “freedom from failure” (FFF),
which included biochemical failure defined as three successive rises in PSA level.27 With a
median follow-up of 60 months, FFF in the 78 Gy arm was 70% compared to 64% in the 70 Gy arm, representing a significant difference (p=0.03). The higher dose was associated with a significantly greater risk of grade 2 or higher late rectal toxicity (26% for 78 Gy versus 12% for 70 Gy; p=0.001). This study was performed before intensity-modulated radiotherapy and other more sophisticated computerized treatment planning were available, and the results for patients with T3 disease could not be separated from those with clinical stage T1 to T2 disease.
A similar French study, the Groupe d’Etude des Tumeurs Uro-Genitales (GETUG) (Table 2),9
reported early toxicity results on 306 patients with clinical stage T1 (Gleason score ≥7 or PSA
≥10 ng/mL) or T2 to T3a disease randomized between 1999 and 2002 to 70 versus 80 Gy. Data
regarding treatment efficacy is not yet available, but the authors reported no significant
differences in treatment toxicity between the two radiation groups. Again, patients with clinical stage T1 to T2 disease were not separable from those with T3a disease.
A multicenter RCT from Loma Linda and Massachusetts General Hospitals (Table 2)35 reported
results for 392 patients with clinical stage T1 to T2 prostate cancer randomized to 70.2 or 79.2
Gy, using a combination of photon and proton beams.35 At five years, there was no difference in
overall survival, but the higher-dose therapy conferred a 49% reduction in the risk of biochemical failure (p<0.001). There was no difference in the incidence of acute or late gastrointestinal or genitourinary toxicity of grade 3 or higher between these two groups. Still, both acute and late grade 2 gastrointestinal toxicity was significantly more common in the high- dose arm.
External Beam Radiotherapy Fractionation
One RCT has reported on efficacy of hypofractionation of external beam radiotherapy and one
study is ongoing. The first, a multicenter Canadian study (Table 2)25 that accrued 936 patients
from 1995 to 1998, randomized men with clinical stage T1 to T2 prostate cancer to 66 Gy in 33 fractions versus 52.5 Gy in 20 fractions. The primary endpoint was biochemical and/or clinical failure, defined as three successive increases in PSA levels, clinical evidence of local or
metastatic failure, commencement of hormonal therapy, or death due to prostate cancer. With a median follow-up of 5.7 years, there was no conclusive evidence for superior efficacy of either treatment regimen. Acute gastrointestinal toxicity was slightly higher in the hypofractionated arm, but there is no difference in late toxicity between the two arms. A similar RCT currently is under way in Australia with comparable findings regarding toxicity, but for which efficacy data are not yet available.34
The Role of Combined Therapy
Neoadjuvant Hormonal Therapy in Combination with Radical Prostatectomy
Several studies have assessed the value of neoadjuvant hormonal therapy (NHT) prior to radical prostatectomy. However, the optimal duration of treatment and the value of this intervention are not yet entirely clear. Initial results from various trials demonstrated a decrease in the rates of positive surgical margins in those men treated with NHT prior to surgery. In a study randomizing 213 men with clinical stage T1b to T2c prostate cancer to radical prostatectomy versus a 12- week course of 300 mg cyproterone acetate with subsequent surgery, Goldenberg et al. (Table
3)16 found positive surgical margins in 64.8% of men undergoing surgery only compared to a
27.7% positive surgical margin rate in the NHT group (p=0.001). While several other groups have reached similar conclusions regarding immediate pathologic outcomes with various NHT combinations and duration,8, 12, 15, 22, 28, 31, 32, 120, 121 it appears that NHT prior to radical
prostatectomy does not impart an overall advantage in terms of biochemical recurrence rates compared to radical prostatectomy alone.8, 21, 31, 32, 120, 121 These findings do not support the routine use of NHT prior to radical prostatectomy.
Hormonal Therapy in Combination with Radiation Therapy
In contrast to the findings of RCTs in the neoadjuvant setting, RCTs studying primary external beam radiotherapy alone or in combination with ADT have demonstrated advantages for radiation and hormonal therapy. In an RCT of 456 men, Radiation Therapy Oncology Group
(Table 4)26 8610 demonstrated improved local control (p=0.016), time to distant metastasis
(p=0.04), and cause-specific survival (p=0.05) for patients with cT2 to T4. In a subset analysis, there was a suggestion that the benefit may be seen more in patients with Gleason score of 6 or lower. Standard external beam radiotherapy with concurrent hormonal ablation that was continued for three years imparts an overall survival advantage (five-year estimates 78% vs. 62%, p=0.0002) among prostate cancer patients with clinical stage T1 to T2 with World Health Organization grade 3 tumors, or cT3 to T4N0-1M0 any grade tumors compared to radiotherapy
alone.11 Similar results have been found by Radiation Therapy Oncology Groups 8531 (Table
4)24 and 9202 (Table 4).17
More recently, D’Amico et al. (Table 4)44 reported the outcomes of 206 men with clinical stage
T1b to T2bNx, PSA levels ≥10 ng/mL, or Gleason score ≥7 who were randomized to six months
of androgen suppression in combination with external beam radiotherapy or radiotherapy alone. All patients were treated with 70 Gy three-dimensional conformal radiotherapy. Those in the combination arm started radiation after two months of treatment with hormonal therapy. This study demonstrated improved disease-specific (p=0.02) and overall survival (p=0.04) in the combined treatment arm with a median follow-up of 4.5 years. In addition, fewer patients required treatment for recurrence in the combination arm (p=0.002).
Other studies have aimed to define the optimal duration and timing of androgen ablation in
combination with radiotherapy. Radiation Therapy Oncology Group 9413 (Table 4)29 was a
randomized 2 x 2 factorial clinical trial designed to test whether whole pelvic (WP) radiotherapy improved progression-free survival compared to prostate-only (PO) radiotherapy and whether neoadjuvant and concurrent hormonal therapy (NCHT) improved progression-free survival compared to adjuvant hormonal therapy in men receiving radiotherapy. Patients treated with WP radiotherapy had superior progression-free survival compared to PO radiotherapy (p=0.02). There was no difference in progression-free survival between the two hormonal treatment
statistical interaction between them. In this study, there appears to be a biologic interaction between the volume radiated and timing of hormonal treatment (p=0.011 for progression-free survival). Essentially, this means that it is more appropriate for this study to be analyzed and reported as a four-arm trial. The investigators note that NCHT was beneficial in terms of progression-free survival for those receiving WP radiotherapy while the adjuvant hormonal
therapy group had more favorable progression-free survival among those with PO radiotherapy.29
Another recently published RCT of 378 men with clinical stage T1c to T4 disease (Table 4)13
suggests that there was no advantage of eight compared to three months of NHT prior to 66 Gy radiotherapy for men with localized prostate cancer. The five-year biochemical failure-free
survival rates were 62% versus 61%, respectively (p=0.36).13 Another smaller clinical trial from
Canada (Table 4)23 found no biochemical-free survival advantage with the addition of adjuvant
hormonal ablation (n=55) versus neoadjuvant hormonal ablation (n=63) and standard
radiotherapy (seven-year estimates of 69% versus 66%, respectively; p=0.60) in a mixed patient population consisting primarily of T2 but also some T3 prostate cancer patients. However, when the sample size is so small, the risk of false-positive and false-negative results is a serious concern.
In summary, many effective therapies for prostate cancer have been developed over time, but there is a paucity of high-quality evidence to favor particular treatment modalities for men with localized prostate cancer, and this evidence is not easily developed. Two examples of the latter phenomenon include the Southwest Oncology Group (SWOG) Study 8890 and the Surgical
ProstatectomyVersus Interstitial Radiation Intervention Trial (SPIRIT). SWOG 8890 attempted
to compare radical prostatectomy to external beam radiotherapy with a goal of randomizing 900 to 1,000 patients. The study accrued a total of six patients in 21 months and was thereafter closed. The same accrual problem occurred with SPIRIT, an RCT comparing radical
prostatectomy with permanent interstitial prostate brachytherapy in patients with clinical stage T1c or T2a disease. Despite considerable efforts and resources to recruit patients, including attempts to enroll patients in the United Kingdom, the study accrued only 56 of the total of 1,980 needed and ultimately closed within 17 months after it was initiated. From these experiences, it seems likely that some trials will never be done due in part to patient and/or physician biases.